CODEWITHSUNDEEP
rmathematical-optimizationquantitative-finance
eadains
eadains

Reputation: 43

PortfolioAnalytics 'from' cannot be NA, NaN, or infinite

Hi I'm getting this error trying to optimize a portfolio for minimum variance:

Error in seq.default(from = round(min, rounding), to = round(max, rounding),  : 
  'from' cannot be NA, NaN or infinite

The optimization runs fine using ROI, simplex, and grid, but all of those methods ignore the position constraint. Same error on random and DEoptim. This is the exact code I'm running:

tickers = c("AMJ", "AMLP", "ASHR", "DIA", "DUST", "DXJ", "EEM", "EFA", 
            "EPI", "ERX", "EWG", "EWI", "EWJ", "EWU", "EWW", "EWY", "EWZ", 
            "FAS", "FAZ", "FEZ", "FXE", "FXI", "FXY", "GDX", "GDXJ", "GLD", 
            "GREK", "HEDJ", "HYG", "IAU", "IBB", "ITB", "IWM", "IYR", "JDST", 
            "JNUG", "KRE", "MDY", "NUGT", "OIH", "QQQ", "SCO", "SDS", "SLV", 
            "SPY", "SSO", "SVXY", "TBT", "TLT", "TNA", "TQQQ", "TZA", "UCO", 
            "UNG", "UPRO", "USO", "UUP", "UVXY", "VIXY", "VWO", "VXX", "XBI", 
            "XHB", "XLB", "XLE", "XLF", "XLI", "XLK", "XLP", "XLU", "XLV", 
            "XLY", "XME", "XOP", "XRT")
data = lapply(tickers, function(sym) {
  getSymbols(sym, src="google", auto.assign=FALSE)
})
close = lapply(data, function(frame) {
  frame[,4]
})
merged = do.call(merge, close)
returns = CalculateReturns(merged)
returns = na.omit(returns)

portfolio = portfolio.spec(assets=colnames(returns))
portfolio = add.constraint(portfolio = portfolio, type="leverage", min_sum=0.99, max_sum=1.01)
portfolio = add.constraint(portfolio, type="position_limit", max_pos=8)
minvar = add.objective(portfolio, type="risk", name="var")
opt = optimize.portfolio(returns, minvar, optimize_method="random")

Here are currently loaded packages:

R version 3.3.3 (2017-03-06)
Platform: x86_64-w64-mingw32/x64 (64-bit)
Running under: Windows >= 8 x64 (build 9200)

locale:
[1] LC_COLLATE=English_United States.1252  LC_CTYPE=English_United States.1252    LC_MONETARY=English_United States.1252
[4] LC_NUMERIC=C                           LC_TIME=English_United States.1252    

attached base packages:
[1] parallel  stats     graphics  grDevices utils     datasets  methods   base     

other attached packages:
 [1] MASS_7.3-45                   quadprog_1.5-5                Rglpk_0.6-3                  
 [4] slam_0.1-40                   iterators_1.0.8               ROI.plugin.quadprog_0.2-5    
 [7] ROI.plugin.glpk_0.2-5         ROI_0.2-6                     DEoptim_2.2-4                
[10] readr_1.1.1                   quantmod_0.4-8                TTR_0.23-1                   
[13] PortfolioAnalytics_1.0.3636   foreach_1.4.3                 PerformanceAnalytics_1.4.3541
[16] xts_0.9-7                     zoo_1.8-0                    

loaded via a namespace (and not attached):
[1] Rcpp_0.12.10     lattice_0.20-34  codetools_0.2-15 grid_3.3.3       R6_2.2.1         registry_0.3     tools_3.3.3     
[8] hms_0.3          tibble_1.3.0

Upvotes: 1

Views: 981

Answers (1)

Lstat
Lstat

Reputation: 1460

It seems you need to set bounds on the weights of the assets. This can be done by adding type="box" constraint. I have rewritten the code a little and removed minvar variable.

portfolio <- portfolio.spec(assets=colnames(returns))
portfolio <- add.constraint(portfolio = portfolio, type="leverage",
              min_sum=0.99, max_sum=1.01)
portfolio <- add.constraint(portfolio, type="position_limit", max_pos=8)
portfolio <- add.constraint(portfolio, type="box", min=0, max=0.5) # <-------
portfolio <- add.objective(portfolio, type="risk", name="var")
opt <- optimize.portfolio(R=returns, portfolio=portfolio, optimize_method="random")

opt

Upvotes: 2

Related Questions